The epidemiology of breast cancer suggests a central etiologic role for premenopausal endogenous sex hormones, yet current data do not indicate which specific endogenous hormones (or hormone fractions) are of greatest importance and at what levels risk is increased. Similarly, current data suggest that urinary estrogen metabolites and plasma antioxidants may also be related to breast cancer risk yet these relations have not been assessed in a large prospective study. The Nurses' Health Study Il (NHSII) is a large ongoing prospective cohort study of 116,678 female, U.S. registered nurses who were 25-42 years Of age when the study began in 1989. We now propose to collect and archive blood and urine samples from approximately 40,000 premenopausal women participating in the NHSII who are at least 35 years of age, are free from cancer, have completed previous NHSII questionnaires, are not using exogenous hormones and are neither pregnant nor breastfeeding. Plasma drawn during both the follicular phase (days 3-5) and luteal phase (approx days 20-22), white and red blood cells, and a urine sample will be stored in liquid nitrogen freezers and later analyzed using a nested case-control design. We propose to assess the following specific hypotheses: (1) higher levels of estrogens, androgens, prolactin and progesterone in both the follicular and luteal phase of the menstrual cycle each increase risk of breast cancer, (2) higher levels of plasma beta-carotene and other carotenoids, retinol, and alpha-tocopherol reduce the risk of breast cancer and (3) the ratio of 16-alpha to 2-hydroxylated urinary metabolites is positively associated with an increased risk of breast cancer. The ongoing NHSII will provide follow-up of the cohort and documentation of breast cancer (CA 50385) in addition to information on important covariates (such as body mass index, parity, age at menarche, smoking status and dietary intake, among others) for the proposed study. Overall, the large size of the cohort, the prospective design, the high follow-up rate, the detailed covariate data, and the ability to collect blood specimens timed according to the menstrual cycle provide a unique opportunity to evaluate several important hypotheses related to breast cancer risk.